How Stars are BornA.C. Danks and P.A. Shaver

There is a vivid interest among astronomers in theearly phases of star formation. In the last issue ofthe Messenger (No. 11, p. 14) a catalogue of stellarbirth places was introduced. The present articlediscusses radio, infrared and optical observationsof a particularly interesting object. The authors areDrs. Anthony C. Danks (ESO-Chile) and PeterA. Shaver (Kapteyn Astronomicallnstitute, Univer­sity of Groningen, the Netherlands).

In recent years both radio and infrared astronomy have re­vealed details of the dusty environment of star formation.These regions are characterized by the presence of "Com­pact H 11 regions" (compact, bright radio continuum sources- Mezger et a/., 1967), wh ich are often associated with H2 0and type lOH masers (showing 1665 and 1667 MHz emission- Habing et a/., 1972). Infrared sources (1 to 30 I-lm) areoften seen in or nearby the compact H 11 regions and some­times combinations of these sources can be found close tovisible H 11 regions.

The "Cocoon" Model

These regions can best be explained quantitatively by therecent models of Kahn (1974) and Cochran and Ostriker(1977), who propose the following scenario: A protostar(M = 40 M0 ) forms by accretion in a dusty interstellar cloud.As the star's luminosity increases with time, the accretion ishalted by radiation pressure. A dusty "cocoon" remains,within which is a smaller ionized zone surrounding the star.In this phase the dust and gas are competing for stellar pho­tons and a situation can arise where the dust is heated to ahigher temperature than the surrounding gas and can giverise to the necessary infrared radiation capable of pumpingthe H2 0 and OH masers. As the star evolves the cocoonfragments and the compact H 11 region become visible. At alater stage, as the star settles into the Main Sequence andthe dust shell dissipates further, a conventional Strömgrensphere (H 11 region) may become visible. This later stage maybe exemplified by regions such as S888 (Pipher et a/., 1977)or Sharpless 2-106 (Sibille etai., 1975) where visible Haemission is seen. Here the infrared source may be inter­preted as an 0 star with high visual extinction due to dust.

bright regions are indicated by 8 and C. Although the Wes­terbork beam is elongated at this low declination, the radiocontours can still be seen to trace out a shell-like structure.We have marked also the positions of the OH sou rce (Evans,private communication) and H2 0 sources (Genzel andDownes, 1978).

Subsequent mapping of the region at 2.21-l at La Sillausing a 10 arcsec diaphragm and 37 arcsec chop on the skyrevealed 3 infrared sources. Two are shown in Figure 1, in­dicated as lAS 1 and 2; the third was detected in the refe­rence beam and is just outside Figure 1. Of these sources,lAS 1 is the most interesting, coinciding with the compactH 11 region. Aecent position measurements of the H2 0source by Jack Welch and Mel Wright using the Hat CreekInterferometer put component A, lAS 1, and the H2 0 sourcewithin 2 arcsec or 0.03 pc- an unusually close association(the source distance of 3.7 kpc was estimated from theH11 Da, H2CO, OH, and H2 0 radial velocities).

We have measured the spectrum of lAS 1 from 1 to 51-l andthis is shown in Figure 2. The upper line in Figure 2 repre-

RIGHT ASCENSION (1950)

Fig. 1. - 6 cm map of G12.2-D.1. The hatched ellipse shows thehalf-power beamwidth. Three continuum sources are indicated asA, Band C. The positions of the OH, H2 0 and infrared sources areshown as crosses.

Observations at Westerbork and La Silla

To investigate the various phases and accompanying phys­ics of these regions, the authors have instigated a pro­gramme to study regions of star formation using the ESO in­frared equipment at the 1 mESO telescope at La Silla.

First results of this programme are shown in Figures 1and 2. In Figure 1 the radio brightness contours are shownfor the sou rce G12.2-0.1. These observations were madewith the Westerbork Synthesis Aadio Telescope at a wave­length of 6 cm; the beam size was 6x31 arcseconds. Thecompact H 11 region is indicated by A and two other radio

18h 09m 50s

f.. 6em

18h 09m 405

17

References

Iy 0 star would be capable of ionizing an H 11 region with100 times the flux of the observed radio source G12.2-0.1.If the star were of spectral type 05 or later then most of thenear-infrared emission could arise from a hot circumstel­lar cocoon (> 1000 K), and even an 07 star would still becapable of powering the radio source. The close H2 0 maserassociation further strengthens our belief that IRS 1 is acocoon star. Further, longer wavelength infrared observa­tions are planned in the coming season in order to deter­mine the spectral type of IRS 1.

Sources of this nature are often associated with large mo­lecular complexes, and 13CO observations have recentlybeen carried out by Michael Scholtes at McDonald. 13COemission has been detected, and 12CO observations will bemade in the near future.

Cochran, W.D., Ostriker, J.P., 1977, Ap. J. 211,392.Genzel, R., Downes, 0., 1977, A&A. Supp/., 30, 145.Habing, H.J., Israel, F.P., de Jong, T., 1972, A&A 17,329.Kahn, F.D., 1974, A&A 37, 149.Mezger, P.G., Altenhoff, W., Schraml, J., Burke, B.F., Reifen­

stein, E.C., Wilson, T.L. 1967. Ap. J. 150, L157.Pipher, J.L., Sharpless, S., Savedoff, M.P., Krassner, J., Varlese,

S., Soifer, B.T., Zeilik 11, M. 1977 A&A 59, 215.Sibille, F., Bergeat, J., Lunel, M., Kandel, R. 1975. A&A 40,441.

sents the total integrated radio flux densities for the entireregion. The infrared spectrum of IRS 1 appears to be stellar,an 03-05 type star with Av = 23 mag. However such an ear-

Fig. 2. - Continuum spectrum of the entire G12.2-0.1 camp/ex(top) and tor component A and /RS 1 (bottam).

-~~:~

~-2

>-IRS1{ \

....,A

> jtilf)

10-1 /1 \10-2 .LL-r,

108 109 10 10 1013 1014 1015

A Magie Eye for Astronomieal SpeetrophotometryThe ability of a telescope to detect faint celestial objects not only depends on the linear size of the telescope, but also uponthe efficiency of the light detectors that are used to register the light. For many years, most astronomical spectra were ob­tained on photographic plates. However, even the best of these rarely achieve detective quantum efficiencies above a fewper cent, i. e. they only "catch" two or three out of every one hundred photons hitting the emulsion. Ouring the past decademuch effort has therefore been concentrated in astronomy on how to impro ve the detector efficiency in order to make sma"telescopes "Iarger" and large telescopes "very large". For instance, a telescope with a mirror diameter of one metre and adetector efficiency of 50 per cent is (for many astronomical applications) equivalent to a 5 metre telescope with a 2 per centdetector.

In this article, ESO engineer Rudi Zurbuchen from the Geneva group discusses one of the new detectors, the RETICONarray.

New Detectors In Astronomy The RETICON Diode Array

Times when astronomers forgot their numb fingers, whilst gazingthrough the eyepiece of a telescope and admiring celestial objectsare definitely over. Today's astronomy and the use of its large opti­cal telescopes require less subjective and much more powerfuleyes. In many astronomical applications electronic detectors aremore and more taking over from the photographic plate. One ofthem, planned to be used with the instruments of the ESO3.6 metre telescope, is described here. The actual hardware andsoftware system is presently being developed bya team of ESO'sInstrument Development Group and will be the subject of a subse­quent article.

A large amount of significant astronomical information such asphysical state, material composition and radial velocity of a stellarobject is retrieved from the precise measurement of the object'sspectrum. The light levels associated with spectrophotometricmeasurements on a good observing site can be very low and therequirements imposed upon efficient light detectors used in thisfield are accordingly high.

The widely-used singel-channel scanning mode of conventionalspectrometers suffers badly from a poor detection efficiencywhich is partly due to the high light loss inherent to the samplingprinciple but also to the modest quantum efficiency of even mod­ern photon multiplier tubes. An additional disadvantage of thesingle-channel scanner is its sensitivity to atmospheric variations.

Among the flood of newly-developed electronic photodetectorsthere is one which is particularly altractive for spectrophotometricapplications.lt is a self-scanned linear photodiode array manufac­tured by the RETICON Corporation, Sunnyvale, California. Severalother array devices are potentially good competitors but the RE­TICON seems, at least for the time being, to be the only one whichprovides as weil a diode sufficiently large to cover a typical as­tronomical spectrum image over its total height, as an adequatelinear field and thereby spectral range.

Reticon linear arrays are available with up to 1872 individualphotodiodes with centre-to-centre spacings as small as 151lm.The first RETICON which will be used for the 3.6 m telescope in­strumentation programme is a dual 1024-element array with a251lm centre-to-centre spacing and an active aperture width of430 11m. Thedual configuration allows simultaneous integration ofobject and background signals and will be used as a near infrareddetector for the low-dispersion spectrograph of the 3.6 m tele­scope Cassegrain focus. Another similar array is planned to be op­erated on the coude echelle high-dispersion spectrometer (see ar­ticle by D. Enard in Messenger No. 11, December 1977).

The RETICON is a monolithic integrated circuit and as such ex­hibits excellent geometric accuracy and stability. Besides thephotodiodes, the circuit has integrated into the same silicon chipthe analog switching circuitry needed for reading out the diode

18